SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2204.00928 · ECCV · 2022

SinNeRF: Training Neural Radiance Fields on Complex Scenes from a Single Image

Zhangyang Wang, Dejia Xu, Yifan Jiang, Peihao Wang, Zhiwen Fan, Humphrey Shi

arXiv · PDF · Open in the Atlas

Code that ran

19 of 22 samples (functions and classes) lifted from this paper's linked repositories executed on a synthesized input in Syntology's jail (0 of them constructed an object rather than computing a result; 3 honoured the contract we drafted, 1 violated it and 11 have no contract verdict; 4 failures of our instrument; 3 no recorded run), measured 2026-11-22; repository links from the Papers with Code archive attributed under CC BY-SA, no affiliation; bodies inline only under a permissive licence as GitHub detects it. Not a reproduction of the paper's results.

"Ran" means the function executed on a synthesized input and returned a value. It is not a reproduction of the paper's results. This record, dated, as JSON: /api/ran/2204.00928.json · badge: /api/ran/2204.00928.svg

Each repository's role is a label: a classifier's output recorded on the link, or, where no role is recorded, what the Papers with Code archive marked. The line under it names its basis. Neither says who wrote the code.

RepositoryRoleRan
VITA-Group/SinNeRF canonical (classifier label)
classifier basis not recorded on the link
19 ran · 3 no recorded run
FunctionStatusWhere it lives
DiffAugment Ran
ran, violated the contract we drafted
VITA-Group/SinNeRF/models/diff_aug.py
code served (permissive licence) · get_code("d39b74f30fd860ba")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned Tensor, ndim 4, dtype family float, shape (2, 3, 4, 5) as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch
import numpy as np

torch.manual_seed(0)
np.random.seed(0)

# Create a small batch of image tensors in channels-first format (B, C, H, W)
x = torch.randn(2, 3, 4, 5, dtype=torch.float32)

ARGS = {
    'x': x,
    'policy': 'color,cutout',
    'channels_first': True,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Tensor, ndim 4, dtype family float, shape (2, 3, 4, 5) — shape (2, 3, 4, 5) as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
actvn Ran
ran; our contract draft was wrong, not the code
VITA-Group/SinNeRF/models/discriminator.py
code served (permissive licence) · get_code("51e3b2175da36d6c")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned Tensor, ndim 2, dtype family float, shape (2, 4) as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch
import torch.nn.functional as F

torch.manual_seed(0)

# The function actvn takes a tensor x and applies leaky_relu with negative slope 0.2
# It needs a tensor as input
x = torch.randn(2, 4)

ARGS = {
    "x": x
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Tensor, ndim 2, dtype family float, shape (2, 4) — shape (2, 4) as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
attn_cosine_sim Ran
ran; our contract draft was wrong, not the code
VITA-Group/SinNeRF/models/extractor.py
pointer only (licence: MIT); the stored copy get_code serves, from omerbt/Splice, is pointer only (licence: NONE) · get_code("e9f612ce15936ed2")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (MIT), but get_code serves the stored copy, first harvested from omerbt/Splice, and GitHub detected no licence in that repository when it was checked. The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. A run is recorded.

create_window Ran
ran on a synthesized input
VITA-Group/SinNeRF/losses.py
code served (permissive licence) · get_code("e83c4d7c9ee62275")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned Tensor, ndim 4, dtype family float, shape (3, 1, 5, 5) as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch
from math import exp

torch.manual_seed(0)

def gaussian(window_size, sigma):
    gauss = torch.Tensor([exp(-(x - window_size // 2) ** 2 / float(2 * sigma ** 2)) for x in range(window_size)])
    return gauss / gauss.sum()

def create_window(window_size, channel):
    _1D_window = gaussian(window_size, 1.5).unsqueeze(1)
    _2D_window = _1D_window.mm(_1D_window.t()).float().unsqueeze(0).unsqueeze(0)
    window = torch.Tensor(_2D_window.expand(channel, 1, window_size, window_size).contiguous())
    return window

ARGS = {
    "window_size": 5,
    "channel": 3
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Tensor, ndim 4, dtype family float, shape (3, 1, 5, 5) — shape (3, 1, 5, 5) as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
eval_points Ran
ran on a synthesized input
VITA-Group/SinNeRF/models/rendering.py
code served (permissive licence) · get_code("c908d8bf570a1296")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned Tensor, ndim 2, dtype family float, shape (2, 1) as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch
import torch.nn as nn

torch.manual_seed(0)

class SimpleEmbedding(nn.Module):
    def __init__(self, input_dim, output_dim):
        super().__init__()
        self.linear = nn.Linear(input_dim, output_dim)
    
    def forward(self, x):
        return self.linear(x)

class SimpleNeRFModel(nn.Module):
    def __init__(self, input_dim, output_dim):
        super().__init__()
        self.linear = nn.Linear(input_dim, output_dim)
    
    def forward(self, x, sigma_only=True):
        return self.linear(x)

# Create points: (N_rays=2, 3)
points = torch.randn(2, 3)

# Create embedding for xyz
embedding_xyz = SimpleEmbedding(input_dim=3, output_dim=4)

# Create fine model
# The model takes embedded xyz as input and outputs sigma values
model_fine = SimpleNeRFModel(input_dim=4, output_dim=1)

# models list with fine model at the end
models = [model_fine]

# embeddings list with xyz embedding at index 0
embeddings = [embedding_xyz]

ARGS = {
    "points": points,
    "models": models,
    "embeddings": embeddings
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Tensor, ndim 2, dtype family float, shape (2, 1) — shape (2, 1) as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
gaussian Ran
ran, honoured the contract we drafted
VITA-Group/SinNeRF/losses.py
code served (permissive licence) · get_code("c56b7ef16f309a45")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1) and returned Tensor, ndim 1, dtype family float, shape (5,) as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch
from math import exp

torch.manual_seed(0)

ARGS = {
    "window_size": 5,
    "sigma": 1.0,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Tensor, ndim 1, dtype family float, shape (5,) — shape (5,) as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
mse Ran
ran on a synthesized input
VITA-Group/SinNeRF/metrics.py
code served (permissive licence) · get_code("c832923b546e9fb3")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned Tensor, ndim 0, dtype family float, shape () as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch

torch.manual_seed(0)

# Create small test tensors
# image_pred: predicted image tensor
image_pred = torch.randn(2, 4)

# image_gt: ground truth image tensor (same shape as image_pred)
image_gt = torch.randn(2, 4)

# valid_mask: boolean mask indicating which elements are valid (optional)
valid_mask = torch.tensor([[True, False, True, False], [True, True, False, True]])

# reduction: string parameter for how to reduce the output
reduction = 'mean'

ARGS = {
    'image_pred': image_pred,
    'image_gt': image_gt,
    'valid_mask': valid_mask,
    'reduction': reduction
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Tensor, ndim 0, dtype family float, shape () — shape () as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
psnr Ran
ran on a synthesized input
VITA-Group/SinNeRF/metrics.py
code served (permissive licence) · get_code("b5b388448f22219a")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned Tensor, ndim 0, dtype family float, shape () as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch

torch.manual_seed(0)

# Create small test tensors
# image_pred: predicted image tensor
image_pred = torch.rand(2, 4, 3, 5)

# image_gt: ground truth image tensor (same shape as image_pred)
image_gt = torch.rand(2, 4, 3, 5)

# valid_mask: optional boolean mask to mask out invalid pixels
valid_mask = torch.ones(2, 4, 3, 5, dtype=torch.bool)
valid_mask[0, 0, 0, 0] = False  # Mark one pixel as invalid

# reduction: how to reduce the result
reduction = 'mean'

ARGS = {
    'image_pred': image_pred,
    'image_gt': image_gt,
    'valid_mask': valid_mask,
    'reduction': reduction,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Tensor, ndim 0, dtype family float, shape () — shape () as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
rand_brightness Ran
ran; our contract draft was wrong, not the code
VITA-Group/SinNeRF/models/diff_aug.py
pointer only (licence: MIT); the stored copy get_code serves, from postech-cvlab/pytorch-studiogan, is pointer only (licence: NOASSERTION) · get_code("011230b2b9b8fb6f")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (MIT), but get_code serves the stored copy, first harvested from postech-cvlab/pytorch-studiogan, and GitHub could not classify the licence of that repository (NOASSERTION). The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. A run is recorded.

rand_saturation Ran
ran; our contract draft was wrong, not the code
VITA-Group/SinNeRF/models/diff_aug.py
pointer only (licence: MIT); the stored copy get_code serves, from postech-cvlab/pytorch-studiogan, is pointer only (licence: NOASSERTION) · get_code("5b0d8787e670fc63")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (MIT), but get_code serves the stored copy, first harvested from postech-cvlab/pytorch-studiogan, and GitHub could not classify the licence of that repository (NOASSERTION). The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. A run is recorded.

render_rays Ran
ran on a synthesized input
VITA-Group/SinNeRF/models/rendering.py
code served (permissive licence) · get_code("15bab88cd78f4699")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned dict. The image it ran in was not recorded. If your copy returns the same type on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch
import torch.nn as nn

torch.manual_seed(0)

# Create minimal NeRF model that accepts xyzdir_embedded and returns output
class MiniNeRF(nn.Module):
    def __init__(self, input_dim=64, output_dim=4):
        super().__init__()
        self.fc = nn.Linear(input_dim, output_dim)
    
    def forward(self, xyzdir_embedded, sigma_only=False):
        return self.fc(xyzdir_embedded)

# Create minimal embedding module for xyz
class EmbeddingXYZ(nn.Module):
    def __init__(self, output_dim=60):
        super().__init__()
        self.output_dim = output_dim
        self.fc = nn.Linear(3, output_dim)
    
    def forward(self, xyz):
        return self.fc(xyz)

# Create minimal embedding module for direction
class EmbeddingDir(nn.Module):
    def __init__(self, output_dim=4):
        super().__init__()
        self.output_dim = output_dim
        self.fc = nn.Linear(3, output_dim)
    
    def forward(self, direction):
        return self.fc(direction)

# Instantiate models and embeddings
coarse_model = MiniNeRF(input_dim=64, output_dim=4)
fine_model = MiniNeRF(input_dim=64, output_dim=4)
embedding_xyz = EmbeddingXYZ(output_dim=60)
embedding_dir = EmbeddingDir(output_dim=4)

models = [coarse_model, fine_model]
embeddings = [embedding_xyz, embedding_dir]

# Create rays: (N_rays, 3+3+2) = (2, 8)
# Format: [ray_origin(3), ray_direction(3), near(1), far(1)]
N_rays = 2
rays = torch.randn(N_rays, 8)

ARGS = {
    'models': models,
    'embeddings': embeddings,
    'rays': rays,
    'N_samples': 64,
    'use_disp': False,
    'perturb': 0,
    'noise_std': 1,
    'N_importance': 0,
    'chunk': 1024 * 32,
    'white_back': False,
    'test_time': False,
    'detach_coarse': False,
    'noisy_coarse': True,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type dict — a dict output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a dict output the recorded ndim 0 and dtype family float describe that coercion, not the output, and are not stated
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
reparametrize Ran
ran on a synthesized input
VITA-Group/SinNeRF/models/discriminator.py
code served (permissive licence) · get_code("d3daffe86c3a152c")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned Tensor, ndim 2, dtype family float, shape (2, 4) as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch

torch.manual_seed(0)

# Create small tensors for mu and logstd
# mu: mean of the distribution, shape (2, 4) for batch_size=2, latent_dim=4
mu = torch.randn(2, 4)

# logstd: log standard deviation, same shape as mu
logstd = torch.randn(2, 4)

ARGS = {
    "mu": mu,
    "logstd": logstd,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Tensor, ndim 2, dtype family float, shape (2, 4) — shape (2, 4) as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
rot_phi Ran
ran, honoured the contract we drafted
VITA-Group/SinNeRF/datasets/blender_ray_patch_1image_proj.py
code served (permissive licence) · get_code("fea8d3701d9a51c3")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned Tensor, ndim 2, dtype family float, shape (4, 4) as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch
import numpy as np

torch.manual_seed(0)

# The function rot_phi takes a single parameter 'phi' which should be a scalar
# (float or int) representing an angle in radians. The function uses it in
# np.cos(phi) and np.sin(phi), so it needs to be a numeric value.

phi = 0.5  # A simple scalar angle in radians

ARGS = {
    "phi": phi
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Tensor, ndim 2, dtype family float, shape (4, 4) — shape (4, 4) as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
rot_theta Ran
ran, honoured the contract we drafted
VITA-Group/SinNeRF/datasets/blender_ray_patch_1image_proj.py
code served (permissive licence) · get_code("379ebe448f8a377f")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned Tensor, ndim 2, dtype family float, shape (4, 4) as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch
import numpy as np

torch.manual_seed(0)

# The function rot_theta takes a single parameter 'th' (theta)
# which is used in np.cos(th) and np.sin(th) calls
# So it should be a scalar numeric value

th = np.pi / 4  # A reasonable angle value in radians

ARGS = {
    "th": th
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Tensor, ndim 2, dtype family float, shape (4, 4) — shape (4, 4) as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
sample_pdf Ran
ran on a synthesized input
VITA-Group/SinNeRF/models/rendering.py
code served (permissive licence) · get_code("7bd5789a725e4ad5")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned Tensor, ndim 2, dtype family float, shape (2, 3) as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch

torch.manual_seed(0)

# Set up dimensions
N_rays = 2
N_samples_ = 4
N_importance = 3

# bins: (N_rays, N_samples_+1)
bins = torch.linspace(0, 1, N_samples_ + 1).unsqueeze(0).expand(N_rays, -1)
bins = bins + torch.randn(N_rays, N_samples_ + 1) * 0.01  # Add small noise for realism

# weights: (N_rays, N_samples_)
weights = torch.rand(N_rays, N_samples_)

# N_importance: scalar int
N_importance = N_importance

# det: boolean
det = False

# eps: small float
eps = 1e-05

ARGS = {
    "bins": bins,
    "weights": weights,
    "N_importance": N_importance,
    "det": det,
    "eps": eps,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Tensor, ndim 2, dtype family float, shape (2, 3) — shape (2, 3) as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
shifted_softplus Ran
ran on a synthesized input
VITA-Group/SinNeRF/models/activations.py
code served (permissive licence) · get_code("fb8e3572d4ab661e")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned Tensor, ndim 2, dtype family float; no output shape is stated. The image it ran in was not recorded. If your copy returns the same type and ndim on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch
import torch.jit as jit
from torch import Tensor

torch.manual_seed(0)

# Create a small input tensor for the shifted_softplus function
# The function takes a single Tensor argument and performs element-wise operations
x = torch.randn(2, 4)

ARGS = {
    "x": x
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Tensor, ndim 2, dtype family float — no output shape is stated: observe_signature.py could not observe it (import:OSError)
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
trans_t Ran
ran on a synthesized input
VITA-Group/SinNeRF/datasets/blender_ray_patch_1image_proj.py
code served (permissive licence) · get_code("16cca4c69adfd265")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned Tensor, ndim 2, dtype family float, shape (4, 4) as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch

torch.manual_seed(0)

# The function trans_t takes a single parameter 't' which is used as a scalar value
# in a 4x4 transformation matrix. The parameter should be a numeric value (int or float).

ARGS = {
    "t": 2.5
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Tensor, ndim 2, dtype family float, shape (4, 4) — shape (4, 4) as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
visualize_depth Ran
ran on a synthesized input
VITA-Group/SinNeRF/datasets/blender_ray_patch_1image_rot3d.py
code served (permissive licence) · get_code("cf55a02a26d5dc86")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned Image. The image it ran in was not recorded. If your copy returns the same type on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import numpy as np
from PIL import Image
import cv2

# Create a small depth map (H=4, W=3)
depth_map = np.array([
    [1.5, 2.0, 2.5],
    [1.8, 2.2, 2.8],
    [1.6, 2.1, 2.6],
    [1.7, 2.3, 2.9]
], dtype=np.float32)

ARGS = {
    "depth": depth_map,
    "cmap": cv2.COLORMAP_JET
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
PIL, cv2, numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Image — a Image output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a Image output the recorded ndim 3 and dtype family int describe that coercion, not the output, and are not stated
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
widened_sigmoid Ran
ran on a synthesized input
VITA-Group/SinNeRF/models/activations.py
code served (permissive licence) · get_code("02c871f5939d0d45")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned Tensor, ndim 2, dtype family float; no output shape is stated. The image it ran in was not recorded. If your copy returns the same type and ndim on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch
import torch.jit as jit
from torch import Tensor

torch.manual_seed(0)

# Create a small input tensor (batch_size=2, features=4)
x = torch.randn(2, 4)

ARGS = {
    "x": x
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Tensor, ndim 2, dtype family float — no output shape is stated: observe_signature.py could not observe it (import:OSError)
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
forward_warp Not yet run
unverified (harvested, no recorded run)
VITA-Group/SinNeRF/models/sinnerf.py
code served (permissive licence) · get_code("f5b5802d716c721e")
How it ran

no invocation recorded.

project_with_depth Not yet run
unverified (harvested, no recorded run)
VITA-Group/SinNeRF/models/sinnerf.py
code served (permissive licence) · get_code("47665077f5888556")
How it ran

no invocation recorded.

ssim Not yet run
unverified (harvested, no recorded run)
VITA-Group/SinNeRF/metrics.py
code served (permissive licence) · get_code("895ffdb66b75e1b4")
How it ran

no invocation recorded.

Also present in other papers' code

Of 22 samples lifted from this paper's linked repositories, 22 carry a function identity and 0 do not. Of the 22 distinct functions with an identity, 13 are also present in other papers whose code we hold (8 with at least one paper dated earlier, 5 the same month, 11 later) and 9 are only here among the 52,977 papers whose code Syntology holds (as of 2026-09-26).

Only here means: among the 52,977 papers whose code Syntology holds (51,887 with lifted samples, 1,090 through their assemblies only), as of 2026-09-26, no other paper's lifted samples or current assembly bricks contain an identical function body after normalisation.

FunctionAlso present inEarliest-datedOthers named
gaussian 160 other papers
20 dated earlier, 1 the same month, 138 later, 1 same year, order not determinable · 7 only as an assembly brick
1812.11941 (2018-12)
High Quality Monocular Depth Estimation via Transfer Learning
1904.01538 (2019-04)
Spatial Attentive Single-Image Deraining with a High Quality Real Rain Dataset
1907.11628 (2019-07)
Unsupervised Learning for Optical Flow Estimation Using Pyramid Convolution LSTM
and 157 more not named here
rand_brightness 17 other papers
7 dated earlier, 1 the same month, 8 later, 1 same year, order not determinable
1912.04958 (2019-12)
Analyzing and Improving the Image Quality of StyleGAN
2006.10738 (2020-06)
Differentiable Augmentation for Data-Efficient GAN Training
2007.15646 (2020-07)
Rewriting a Deep Generative Model
and 14 more not named here
rand_saturation 16 other papers
7 dated earlier, 8 later, 1 same year, order not determinable
1912.04958 (2019-12)
Analyzing and Improving the Image Quality of StyleGAN
2006.10738 (2020-06)
Differentiable Augmentation for Data-Efficient GAN Training
2007.15646 (2020-07)
Rewriting a Deep Generative Model
and 13 more not named here
actvn 6 other papers
4 dated earlier, 2 later
1810.00821 (2018-10)
Variational Discriminator Bottleneck: Improving Imitation Learning, Inverse RL, and GANs by Constraining Information Flow
1912.05237 (2019-12)
Towards Unsupervised Learning of Generative Models for 3D Controllable Image Synthesis
2002.11810 (2020-02)
On Leveraging Pretrained GANs for Generation with Limited Data
and 3 more not named here
mse 5 other papers
3 dated earlier, 1 the same month, 1 later
2008.02268 (2020-08)
NeRF in the Wild: Neural Radiance Fields for Unconstrained Photo Collections
2106.13228 (2021-06)
HyperNeRF: A Higher-Dimensional Representation for Topologically Varying Neural Radiance Fields
2112.01523 (2021-12)
Learning Neural Light Fields with Ray-Space Embedding Networks
and 2 more not named here
attn_cosine_sim 3 other papers
1 dated earlier, 2 later
2201.00424 (2022-01)
Splicing ViT Features for Semantic Appearance Transfer
2305.19201 (2023-05)
DäRF: Boosting Radiance Fields from Sparse Inputs with Monocular Depth Adaptation
2401.09146 (2024-01)
Continuous Piecewise-Affine Based Motion Model for Image Animation
psnr 3 other papers
2 dated earlier, 1 the same month
2008.02268 (2020-08)
NeRF in the Wild: Neural Radiance Fields for Unconstrained Photo Collections
2106.13228 (2021-06)
HyperNeRF: A Higher-Dimensional Representation for Topologically Varying Neural Radiance Fields
2204.00559 (2022-04)
DFNet: Enhance Absolute Pose Regression with Direct Feature Matching
ssim 3 other papers
2 dated earlier, 1 the same month
2008.02268 (2020-08)
NeRF in the Wild: Neural Radiance Fields for Unconstrained Photo Collections
2106.13228 (2021-06)
HyperNeRF: A Higher-Dimensional Representation for Topologically Varying Neural Radiance Fields
2204.00559 (2022-04)
DFNet: Enhance Absolute Pose Regression with Direct Feature Matching
DiffAugment 1 other paper
1 later
2305.19201 (2023-05)
DäRF: Boosting Radiance Fields from Sparse Inputs with Monocular Depth Adaptation
reparametrize 1 other paper
1 later
2305.19201 (2023-05)
DäRF: Boosting Radiance Fields from Sparse Inputs with Monocular Depth Adaptation
rot_phi 1 other paper
1 later
2301.09632 (2023-01)
HexPlane: A Fast Representation for Dynamic Scenes
rot_theta 1 other paper
1 later
2301.09632 (2023-01)
HexPlane: A Fast Representation for Dynamic Scenes
trans_t 1 other paper
1 later
2301.09632 (2023-01)
HexPlane: A Fast Representation for Dynamic Scenes

Presence, not direction: an identical function body in two papers' repositories does not say who wrote it, which came from which, or that either paper's method rests on it. Matching is exact after normalisation, so a renamed or edited function is not matched. Dates are when papers appeared, not when code was written. Only .py files the harvester lifted are covered; notebooks are not read. A new-style arXiv id YYMM.NNNN dates a paper to the month 20YY-MM; otherwise Paper.year dates it to the year; otherwise it is undated. Dates are when the paper appeared, not when the code was written. Read 2026-11-22.

Repositories linked to this paper

Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.

For agents

The same record, over MCP at https://syntology.ai/mcp:

get_harvested_code_for_paper("2204.00928")
get_code_for_paper("2204.00928")
have("2204.00928")

The run record, dated, one paper per request, free:

curl https://syntology.ai/api/ran/2204.00928.json

A badge for a README (the split and the date, never a ratio):

[![Syntology run record](https://syntology.ai/api/ran/2204.00928.svg)](https://syntology.ai/paper/2204.00928)

Connect an agent — have() is free.